In a continuous kneading apparatus according to an embodiment, for each of a plurality of ring-shaped heaters, a control unit determines a current state and a reward for an action selected in the past based on a control error calculated from an acquired temperature; updates a control condition based on the reward, and determines an optimum action corresponding to the current state under the updated control condition, the control condition being a combination of a state and an action; and controls a target ring-shaped heater based on the optimum action.